Merge branch 'upstream' into concedo_experimental

# Conflicts:
#	.devops/cann.Dockerfile
#	.devops/cpu.Dockerfile
#	.devops/cuda.Dockerfile
#	.devops/intel.Dockerfile
#	.devops/musa.Dockerfile
#	.devops/nix/package.nix
#	.devops/rocm.Dockerfile
#	.devops/vulkan.Dockerfile
#	.github/workflows/build.yml
#	.github/workflows/server.yml
#	CMakeLists.txt
#	build-xcframework.sh
#	ci/run.sh
#	common/CMakeLists.txt
#	common/download.cpp
#	docs/backend/CANN.md
#	docs/ops.md
#	docs/ops/SYCL.csv
#	ggml/src/CMakeLists.txt
#	ggml/src/ggml-cann/aclnn_ops.cpp
#	ggml/src/ggml-cann/aclnn_ops.h
#	ggml/src/ggml-cann/ggml-cann.cpp
#	ggml/src/ggml-cpu/CMakeLists.txt
#	ggml/src/ggml-cpu/kleidiai/kernels.cpp
#	ggml/src/ggml-cpu/kleidiai/kernels.h
#	ggml/src/ggml-cpu/kleidiai/kleidiai.cpp
#	ggml/src/ggml-sycl/ggml-sycl.cpp
#	scripts/sync_vendor.py
#	src/CMakeLists.txt
#	tests/test-backend-ops.cpp
#	tests/test-rope.cpp
#	tools/mtmd/CMakeLists.txt
#	tools/rpc/CMakeLists.txt
#	tools/server/CMakeLists.txt

src/llama-graph.cpp

@@ -1592,9 +1592,10 @@ ggml_tensor * llm_graph_context::build_attn(
             int       il) const {
     // these nodes are added to the graph together so that they are not reordered
     // by doing so, the number of splits in the graph is reduced
+    // expand k later to enable rope fusion which directly writes into k-v cache
     ggml_build_forward_expand(gf, q_cur);
-    ggml_build_forward_expand(gf, k_cur);
     ggml_build_forward_expand(gf, v_cur);
+    ggml_build_forward_expand(gf, k_cur);
 
     const auto * mctx_cur = inp->mctx;
 

src/llama-vocab.cpp

@@ -1238,7 +1238,7 @@ private:
         }
     private:
         uint32_t get_node(size_t index) {
-            if (index > xcda_array_size) {
+            if (index >= xcda_array_size) {
                 throw std::runtime_error("Index out of array bounds in XCDA array!");
             }
             return xcda_array[index];